Reaction time - Relationship between capacity for motor cortical plasticity and learning a comp

8 METHODS

9.1.2 Reaction time

Average reaction time (RT) was 216 ms (SD=0.048, n=13) at baseline and did not change statistically significantly as a group during the intervention (Figure 14). Average change of reaction time from first to fifth training session was -2 ms (SD=21, n=12, p=0.64) and from first to retention training session -4 ms (SD=34, n=13, p=0.46). RT1 correlated significantly

with RTRET (rs=0.61, p=0.03) but not with RT5 (rs=0.36, p=0.26). Reaction time on first session correlated negatively and statistically significantly with change in reaction time ΔRT1–RT5 (n=12, rs=-0.68, p=0.02) and with a trend with ΔRT1–RTRET (n=13, rs=-0.48, p=0.09). Reaction time change ΔRT1–RT5 correlated with ΔRT1–RTRET (n=12, rs=0.685, p=0.01).

FIGURE 14. Reaction time at the start of the first (RT1), fifth (RT5) and retention (RTRET) training sessions.

Correlation between reaction time and juggling skill. Baseline reaction time RT1 correlated significantly with the baseline juggling test result (n=13, rs=-0.59, p=0.03) and the transfer skill test (n=13, rs=-0.59, p=0.04). The change in reaction time from RT1 to RTRET correlated significantly with change of CPA during the same timeline (rs=0.66, p=0.01). The same was not true for the development of the parameters from day 1 to day 5 (n=12, rs=-0.07, p=0.83). Skill acquisition day correlated almost significantly with reaction time change from RT1 to RTRET (n=13, rs=-0.55, p=0.053). Reaction time or reaction time changes did not correlate with relative consolidation POST MT1, relative retention or relative transfer.

0.0 0.1 0.2 0.3

RT1 RT5 RTRET

Reaction time (s)

Reaction Time n=13

n=12 n=13

46 TMS

9.2

9.2.1 MEP amplitudes and MEP changes

PAS session. Average MEP amplitudes increased compared to PRE value in 8 of 13 participants right after PAS and in 8 of 12 participants 20 minutes after PAS. For six participants MEP amplitudes were elevated both right after PAS and 20 minutes after. On average MEP amplitudes increased PRE to POST by 18 % (SD=36, n=13) and PRE to POST20 by 16 % (SD=31, n=12). Differences between MEP amplitudes were not significant PRE to POST (p=0.28, n=13) or PRE to POST20 (p=0.07, n=12) (Figure 15 A). No statistically significant differences were seen between MEP amplitudes at any single stimulus intensities from PRE to POST values or PRE to POST20 values (Figure 15 B). Maximal M-wave amplitude was 9.10 mV (SD=4.62, n=12) PRE and 9.23 mV (SD=4.78, n=12) POST and there was no statistically significant change (p=0.75, n=12).

FIGURE 15. Average MEP amplitudes before (PRE, n=13), right after (POST n=13) and 20 minutes after (POST 20 min, n=12) PAS. There were no statistically significant differences between PRE and POST or POST 20 min MEP amplitudes. Figure A) Shows MEP amplitude averages from all stimulus intensities. There was a trend for increase of average MEP PRE to POST20 (p=0.07). Figure B) shows the Input/Output curve.

MT1. MEP sizes decreased right after first juggling training session for 10 participants and increased for 3 (n=13). After 20 minutes from juggling MEP sizes decreased for 8 participants and increased for 5 participants (n=13) compared to PRE. Average DELTA% right after training was -25 % (SD=32) and -14 % (SD=32) after 20 minutes (n=13). Average MEP size differences were not statistically significant between PRE and POST (n=13, p=0.08) or PRE and POST20 (n=13, p=0.13) (Figure 16 A). When looking at single intensities, the change in MEP sizes was significant with intensities of 110, 120 and 140 %RMT right after MT1 and with intensities of 120 and 140 %RMT 20 min after MT1 (Figure 16 B). Maximal M-wave amplitude was 9.04 mV (SD=3.38, n=11) PRE and 9.02 mV (SD=3.57, n=11) POST and there was no statistically significant change (p=1.00, n=11).

FIGURE 16. Average MEP amplitudes before (PRE), right after (POST) and 20 minutes after (POST 20 min) first motor skill training session (n=13). Figure A) Shows MEP amplitude averages from all stimulus intensities. Figure B) shows Input/Output curve, *=statistically significant difference (p<0.05) compared to PRE value.

MT5. MEPs at fifth juggling session decreased right after training for 10 participants and 20 min after for 8 participants (n=13). Average DELTA% was -12% (SD=39) right after and 4 % (SD=49) twenty minutes after compared to PRE (n=13). Average MEP change from PRE to POST was statistically significant (n=13, p=0.05) but the same was not true from PRE to

difference in MEP sizes was only seen right after juggling training at intensity of 140 %RMT (Figure 17 B). Maximal M-wave amplitude was 8.48 mV (SD=2.66, n=10) PRE and 9.15 mV (SD=3.30, n=10) POST and there was no statistically significant change (p=0.11, n=10).

FIGURE 17. Average MEP amplitudes before (PRE), right after (POST) and 20 minutes after (POST 20 min) fifth motor skill training session (n=13). Figure A) Shows MEP amplitude averages from all stimulus intensities. Figure B) shows Input/Output curve. *=statistically significant difference (p<0.05) compared to PRE value.

Retention Session. MEP amplitudes decreased right after retention training session for 10 participants and 20 min after for 8 participants (n=13). On average MEP sizes changed -25 % (SD=39) right after juggling and -8 % (SD=36) 20 minutes after juggling compared to PRE (n=13). MEP change was statistically significant from PRE to POST (n=13, p=0.03*) but not PRE to POST20 (n=13, p=0.50) (Figure 18 A). At different stimulus intensities a statistically significant difference was found at 110 and 140 %RMT PRE to POST and at 140 %RMT PRE to POST20 (Figure 18 B). Maximal M-wave amplitude was 9.34 mV (SD=3.04, n=11) PRE and 8.99 mV (SD=3.22, n=10) POST and there was no statistically significant change (p=0.37, n=11).

FIGURE 18. Average MEP amplitudes before (PRE), right after (POST) and 20 minutes after (POST 20 min) retention motor skill training session (n=13). Figure A) Shows MEP amplitude averages from all stimulus intensities. Figure B) shows Input/Output curve. *=statistically significant difference (p<0.05) compared to PRE value.

Multi-session baseline MEP amplitude development. Average MEPPRE/Mmax from different days were similar sized and did not change statistically significantly at any point during the training (n=8) (Figure 19). Table shows average changes and standard deviations of baseline MEP changes from PAS to MT1, from MT1 to MT5 and from MT1 to MTRET (Table 6).

FIGURE 19. Normalized baseline MEP amplitudes on PAS, MT1, MT5 and retention sessions (n=8).

0.0

TABLE 6. The percentage change in normalized baseline MEP amplitudes during the study (n=8).

Change in MEP_PRE/M_max (Δ%)

M SD p

PASPRE–MT1PRE 19 55 0.58

MT1PRE–MT5PRE 18 49 0.58

MT1PRE–RETPRE 10 33 0.40

9.2.2 Correlations between MEP amplitude changes

In-session correlations. In-session MEP changes Δ%POST and Δ%POST20 correlated statistically significantly on PAS session (n=12, rs=0.75, p=0.005). In-session MEP changes on MT1 correlated but not statistically significantly (n=13, rs=0.52, p=0.07). In-session MEP changes on MT5 correlated statistically significantly (n=13, r=0.70, p=0.008). In-session MEP changes on retention session did not correlate statistically significantly (n=13, rs=0.48, p=0.10).

Correlation between MEP changes on PAS and MEP changes on later sessions. Δ%PASPOST

correlated with Δ%RETPOST20 (n=13, rs=-0.59, p=0.04). Δ%PASPOST also correlated, though not statistically significantly with Δ%MT5POST20 (n=13, rs=0.43, p=0.14) and Δ%RETPOST

(n=13, rs=-0.45, p=0.12). Other correlations between Δ%PASPOST and MEP changes on other sessions were small and not significant. Δ%PASPOST20 did not correlate with MEP changes during MT1, MT5 or MTRET.

Correlation between MEP changes on MT1 and later session. Δ%MT1POST correlated negatively and significantly with Δ%MT5POST (n=13, rs=-0.63, p=0.02) and near significantly wit Δ%MT5POST20 (n=13, rs=-0.54, p=0.06). There was a negative trend for correlation between Δ%MT1POST20 and Δ%RETPOST20 (n=13, rs=-0.51, p=0.08). No other notable correlations were observed between MEP amplitude changes from different sessions.

Correlation between in-session MEP changes and multi-session baseline MEP amplitude changes. In-session MEP change PRE to POST20 PAS correlated with baseline MEP

amplitude change from PAS to MT1 (Figure 20) but otherwise in-session changes during PAS did not correlate with baseline MEP changes. In-session MEP change Δ%MT1POST correlated statistically significantly with baseline MEP change from PAS to MT1 (rs=-0.81 p=0.02, n=8) and not significantly with baseline MEP changes from MT1 to MT5 (rs=0.62, p=0.10, n=8).

In-session MEP change Δ%MT1POST20 did not correlate with baseline MEP amplitude changes other than with baseline MEP change from PAS to MT1 (rs=-0.71, p=0.05, n=8).

Correlation between in-session MEP changes on later sessions (MT5 and MTRET) and baseline MEP changes were not tested.

FIGURE 20. Scatter diagram of acute MEP amplitude change PRE to 20 min POST PAS and baseline MEP amplitude change from PAS to MT1.

Correlations between MEP changes and results of juggling and reaction time 9.3

Correlations between MEP changes on PAS and motor learning parameters. Change in MEP size from PRE to POST PAS correlated with negatively and statistically significantly with percentage transfer (Figure 21 A), but correlation was no longer significant at Δ%POST20

(Figure 21 B). Otherwise MEP changes did not correlate with juggling results (Table 7). A statistically significant correlation was found between reaction time at the retention and

-80 -40 0 40 80 120

-40 -30 -20 -10 0 10 20 30 40

Δ % Baseline MEP PAS–MT1

Δ%PAS POST20

Scatter Δ%PAS POST20 and Baseline MEP change PAS–MT1 n=7 r=0.86 p=0.01*

Δ%PASPOST20 (n=12, rs=0.58, p=0.05). Otherwise Δ%PAS and RT results did not correlate statistically significantly.

FIGURE 21. Scatter diagram of MEP changes on PAS and relative skill transfer (Δ%). A) MEP change from PRE to POST, and B) MEP change from PRE to 20 min POST. r=Spearman's rho correlation coefficient.

-80 -40 0 40 80

-40 -20 0 20 40 60 80 100

Transfer (Δ %)

Δ%PAS POST

Scatter Δ%PAS POST and Percentage Transfer

n=13 r=-0.67 p=0.01*

-80 -40 0 40 80

-40 -20 0 20 40 60 80 100

Transfer (Δ %)

Δ%PAS POST

Scatter Δ%PAS POST20 and Percentage Transfer n=12 r=-0.42 p=0.18

TABLE 7. Correlation between MEP change during PAS and results of juggling skill and reaction time. Percentage change of MEPs was analysed from PRE to POST (Δ%PASPOST) and from PRE to 20 minutes POST PAS (Δ%PASPOST20). retention of skill (Δ%), TRANSF= percentage transfer of skill (Δ%), RT= reaction time (s), ΔRT=reaction time change (s).

rs= Spearman´s Rho correlation coefficient

*= p<0.05; **=p<0.01

Correlations between MEP changes on MT1 and motor learning parameters. A trend of correlation was found between Δ%MT1POST and percentage transfer (Δ%) (n=13, rs=0.54, p=0.06), but otherwise MEP changes on first training session did not correlate with juggling skill results (Table 8). MEP change from PRE to POST20 correlated statistically significantly with RT5 (Figure 22 A) and near significantly with RTRET (n=13, rs=-0.53, p=0.06).

Participants that increased MEP amplitudes PRE to POST20 also improved their reaction times from RT1 to RT5, but correlation between MEP change and reaction time change was not statistically significant (Figure 22 B).

FIGURE 22. Correlation between MEP change from PRE to 20 minutes POST during first training session and A) reaction time on MT5 and B) reaction time change from first training day to fifth.

r=Spearman's rho correlation coefficient, n=12

n=12 r=-0.62 p=0.03*

0.15 0.30

-80 -60 -40 -20 0 20 40 60

Reaction time (s)

Δ% MT1 POST 20 min Scatter Δ%MT1 POST20 and RT5

n=12 r=-0.36 p=0.26

-0.05 0.00 0.05

-80 -60 -40 -20 0 20 40 60

Reaction time (s)

Δ% MT1 POST 20 min

Scatter Δ%MT1 POST20 and ΔRT1 - RT5

TABLE 8. Correlation between MEP change during first training session and results of juggling skill and reaction time. Percentage change of MEPs was analysed from PRE to POST (Δ%MT1POST) and from PRE to 20 minutes POST (Δ%MT1POST20).

Δ%MT1POST Δ%MT1POST20

Abbreviations: CPA= Catches per attempt of juggling, ΔCPA= Change of juggling skill, Skill acquisition = session when CPA≥4, CONS_MT1=percentage consolidation of skill after MT1 (Δ%), RETΔ%= percentage retention of skill (Δ%), TRANSF= percentage transfer of skill (Δ%), RT= reaction time (s), ΔRT=reaction time change (s).

rs= Spearman´s Rho correlation coefficient

*= p<0.05; **=p<0.01

Correlations between MEP changes on later training sessions and motor learning parameters. MEP changes on MT5 did not correlate either with juggling results or reaction time results (Appendix 2). Δ%RETPOST correlated, though not statistically significantly with reaction time at retention (n=13, r=0.51, p=0.07), but otherwise MEP changes did not correlate with juggling or reaction time results on retention session (Appendix 3).

Correlations between multi-session baseline MEP amplitude changes and motor learning parameters. Percentage changes of baseline MEP/Mmax did not correlate statistically significantly with juggling skill results (Table 9). Percentage change of baseline MEP/Mmax

MT1–MT5 correlated positively with reaction time RT1 (Figure 23 A) and negatively with change of reaction time ΔRT1–RT5 (Figure 23 B).

FIGURE 23. Relationship between multi-session change of baseline MEPs and A) baseline reaction time, B) change of reaction time from first to fifth training session. r=Spearman's rho correlation coefficient, n=8.

n=8 r=0.74 p=0.01*

0.15 0.30

-50 0 50 100 150

Reaction time (s)

Δ% baseline MEPs Scatter Δ%MT1–MT5 and RT1

n=8 r=-0.81 p=0.01*

-0.05 0.00 0.05

-40 -20 0 20 40 60 80 100 120

ΔRT (s)

Δ% baseline MEPs

Scatter Δ%MT1–MT5 and ΔRT1–RT5

TABLE 9. Correlation between multi-session baseline MEP change and results of juggling skill and reaction time. Percentage change of baseline MEPs was calculated from MT1 to MT5 and from MT1 to MTRET. n=8. retention of skill (Δ%), TRANSF= percentage transfer of skill (Δ%), RT= reaction time (s), ΔRT=reaction time change (s).

rs= Spearman´s Rho correlation coefficient

*= p<0.05; **=p<0.01

Effect of recreational activity 9.4

Five participants participated in some activities that exploit novel motor control of hand area including handcrafts (F=1), guitar playing (F=2) and piano playing (F=2). All of them

reported that they where not training actively at the time of the study. All but one participant participated in regular sports or exercise activities (Table 10).

TABLE 10. The number of participants in each recreational activity level category (n=13).

Recreational activity level <1 times/wk 1-2 times/wk 3-5 times/wk > 5 times/wk

SPORTS 1 3 6 3

DEX 8 5 0 0

SPORTS&DEX 1 3 5 4

SPORTS= Activity of sports and exercise, DEX= activity of activities employing manual dexterity, DEX= combined activity of sports, exercise and motor activities employing manual dexterity.

Activity levels and juggling skill. Sports activity level correlated with following juggling results: PRE (n=13, rs=0.59, p=0.04), retention (n=13, rs=0.64, p=0.02), transfer (n=13, rs=0.57, p=0.04) and ΔCPAPRE–RET (n=13, rs=0.64, p=0.02). Participation to activities employing manual dexterity (DEX) correlated near significantly with CPA at retention (n=13, rs=0.55, p=0.05), and with ΔCPAPRE–RET (n=13, rs=0.55, p=0.05), consolidation of MT1 (n=13, rs=0.51, p=0.08) and relative transfer (n=13, rs=0.55, p=0.05). Combined activity level SPORTS&DEX correlated statistically significantly with juggling results PRE (rs=0.68, p=0.01), Retention (r=0.73, p=0.004) and transfer (rs=0.64, p=0.02), ΔCPAPRE–RET (n=13, rs=0.73, p=0.004) and near significantly with MT5POST (rs=0.54, p=0.06) and ΔCPAPRE–

MT5POST (n=13, rs=0.54, p=0.06).

Activity levels and reaction time. Sports activity correlated with ΔRT1–RT5 (n=12, rs=0.59, p=0.05). DEX correlated with reaction time change from PRE to Retention (n=13, rs=0.59, p=0.03). SPORTS&DEX did not correlate with reaction time results. Otherwise activity levels did not correlate with reaction time results.

MEP changes and motor activity. DEX correlated negatively with Δ%PASPOST (rs=-0.51, p=0.08) though not statistically significantly. Otherwise in-session or multi-session MEP amplitude changes did not correlate with recreational activity levels during any motor skill training session.

59 10 DISCUSSION

The purpose of this study was to examine the relationship between capacity for motor cortical plasticity and motor skill learning of a complex perceptual-motor skill. Juggling skill developed statistically significantly during the intervention whereas simple visual reaction time did not change as a group. PAS induced an increase of MEP amplitudes that did not reach a statistical significance. PAS induced increase of corticospinal excitability did not correlate with juggling skill development but was associated with slower reaction times at retention. All juggling training sessions induced an immediate suppression of MEP amplitudes that weakened over time. However, MEP amplitudes increased over baseline during 20-minute period after the end of first juggling training session for five participants who were also among those that improved their visual reaction times during the intervention.

Multiple motor training sessions did not induce significant change in baseline corticospinal excitability. Improvement of reaction time from first to fifth training session was associated with increase of baseline corticospinal excitability during same timeline. Motor training induced acute or multisession corticospinal excitability changes did not correlate with juggling skill development. Acute and multisession corticospinal excitability changes induced by PAS and different juggling sessions correlated in complex ways.

Juggling skill 10.1

All participants got better at the juggling task during the 5-day training period but not all acquired the skill (criteria CPA ≥ 4), which was an expected result on the grounds of earlier studies (Bebko et al 2003; Laughlin et al 2015). High level of long-term retention of learned motor skills has been reported in scientific literature since the early 20^th century (Adams 1987). As anticipated, in the present study the participants demonstrated a long lasting learning effect of the trained juggling skill that was still apparent 6 days after the end of the training period.

During a training period performance of motor skill develops both during the training session and consolidation periods between training sessions (Korman et al 2003). In the present study

the in-session gains of skill were statistically significant on first, second, fourth and retention skill training sessions but not on third and fifth training sessions. Gains of skill were retained after 24 h consolidation periods. Many individuals experienced gains of CPA during consolidation periods of 24 h but the gains were not statistically significant as a group. In conclusion gains of skill were attained mainly during the training sessions.

Higher gain of skill was associated with higher performance level at retention test. Greater improvement in juggling performance during the 5-day training period was associated with greater relative drop of performance during 6-day break from training. Similarly greater gain of skill during the first training session was associated with poorer percentage retention after training intervention. The finding is in line with a theory of cognitive effort: greater cognitive effort should result in either slower or normal learning rate during acquisition but better retention (Lee et al 1994). However, it should be noted, that in the present study the acquired skill level was well preserved and those that gained skill faster also had higher retention and transfer test performance levels. High variation in skill also made comparison of relative gains of performance more difficult, which is why the relative learning rate was not analysed at all.

Therefore comparing relative retention according to learning speed is difficult.

Skill transferred to a modified juggling task where juggling pattern remained same but the three balls were all of different sizes. There was a trend for decrease of successful throws from post retention to transfer that was similar to the decrease of skill level reported in Laughlin et al (2015). In the present study juggling skill learning during the acquisition period did not affect how well the obtained skill level was transferred to modified task. All in all the juggling skill transferred well to a modified juggling task.

Relationship between reaction time and juggling skill 10.2

It is known that some types motor skill training and reaction time task training can improve reaction time and movement response time (e.g. Ando et al 2002; Ando et al 2004; Dartnall et al 2009; Proctor et al 1991). In the present study no statistically significant changes in reaction time were observed. However reaction time was more likely to improve for those that

had slower initial reaction times, which is in line with findings of previous research literature (Yotani 2011).

In the present study faster initial reaction time was associated with higher baseline juggling performance level and transfer test performance. As task response times and reduced errors go hand in hand, it could explain the finding of this study (Dartnall et al 2009). It could be that faster reaction time in itself enhances juggling performance. However history of motor activity could have been an effector as well.

Interestingly improvement in reaction time was inversely related to development of juggling skill from baseline test to retention. In other words, reaction time improved in slower learners during juggling training. Maintaining a successful juggling pattern requires ability to utilize juggling pattern specific temporal, spatial and visual control strategies (Huys et al 2004). If hypothesized that the slow learners did not develop the necessary control strategies it would make sense that the training targeted different aspects of perceptual motor function than it did for fast learners. It is speculative but catching the balls might have become a reaction time task for the slow learners, which could also explain improvement in reaction time. Also reaction time development might be associated with the development of some of the sub-skills of juggling necessary to learning. This study however only measured the acquisition of the whole skill and not the sub-skills which is why such conclusion cannot be reached in basis of this study.

PAS induced plasticity and motor skill learning 10.3

Paired associative stimulation was aimed to induce LTP-like plasticity of motor cortex in flexor carpi radialis (FCR) muscle motor area. A trend for increase of corticospinal excitability was found twenty minutes after PAS. Based on literature (Delvendahl et al 2012) it is likely that responders to PAS stimulation experienced LTP-like motor cortical plasticity in this study. Spinal excitability was not measured but it is possible that PAS may have induced a change in spinal reflex modulation, an effect that has been observed after PAS targeting the FCR muscle (Lamy et al 2010; Meunier et al 2007).

For this study a 20ms interstimulus interval was used as it has been reported to induce a statistically significant elevation of MEP amplitudes in flexor carpi radialis muscle (Lamy et al 2010; Meunier et al 2007). It is known that the inter-individual and intra-individual variation in PAS induced neuroplasticity is high (Fratello et al 2006). In a small sample the high inter-individual variation may prevent the results from reaching statistical significance (López-Alonzo et al 2018). In the present study one participant had to leave prematurely before POST20 measurement of PAS session, making the dataset from PAS incomplete and further reduced the likelihood of the MEP change reaching statistical significance. A number of factors are known to influence PAS effect (e.g. Sale et al 2007; Stefan et al 2004). For example hormonal levels and attention during measurement were not measured in the present study. Also the time of the day for PAS measurement differed between participants because of

In document Relationship between capacity for motor cortical plasticity and learning a complex perceptual-motor skill (sivua 51-95)